Genetic diversity of Capsicum chinensis
(Solanaceae) accessions based on molecular
markers and morphological and agronomic traits
F.L. Finger1, S.D. Lannes1, A.R. Schuelter2, J. Doege2, A.P. Comerlato2,
L.S.A. Gonçalves3, F.R.A. Ferreira4, L.R. Clovis4 and C.A. Scapim4
Departamento de Fitotecnia, Universidade Federal de Viçosa, Viçosa, MG, Brasil
Laboratorio de Biotecnologia, Universidade Paranaense,
Campus Toledo, Toledo, PR, Brasil
3
Laboratorio de Genética e Melhoramento de Plantas,
Universidade Estadual do Norte Fluminense Darcy Ribeiro,
Campos dos Goytacazes, RJ, Brasil
4
Departamento de Agronomia,
Universidade Estadual de Maringá, Maringá, PR, Brasil
1
2
Corresponding author: F.L. Finger
E-mail: ffinger@ufv.br
Genet. Mol. Res. 9 (3): 1852-1864 (2010)
Received May 23, 2010
Accepted June 23, 2010
Published September 21, 2010
DOI 10.4238/vol9-3gmr891
ABSTRACT. We estimated the genetic diversity of 49 accessions of
the hot pepper species Capsicum chinensis through analyses of 12
physicochemical traits of the fruit, eight multi-categorical variables,
and with 32 RAPD primers. Data from the physicochemical traits
were submitted to analysis of variance to estimate the genetic
parameters, and their means were clustered by the Scott-Knott
test. The matrices from the individual and combined distance
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
Genetic diversity of Capsicum chinensis
1853
were estimated by multivariate analyses before applying Tocher’s
optimization method. All physicochemical traits were examined
for genetic variability by analysis of variance. The responses
of these traits showed more contribution from genetic than from
environmental factors, except the percentage of dry biomass,
content of soluble solids and vitamin C level. Total capsaicin had
the greatest genetic divergence. Nine clusters were formed from the
quantitative data based on the generalized distance of Mahalanobis,
using Tocher’s method; four were formed from the multi-categorical
data using the Cole-Rodgers coefficient, and eight were formed
from the molecular data using the Nei and Li coefficient. The
accessions were distributed into 14 groups using Tocher’s method,
and no significant correlation between pungency and origin was
detected. Uni- and multivariate analyses permitted the identification
of marked genetic diversity and fruit attributes capable of being
improved through breeding programs.
Key words: Physicochemical traits; Multi-categorical variables;
Multivariate analysis; Capsicum chinensis
INTRODUCTION
The Capsicum genus contains numerous species of sweet and hot peppers (Reifschneider, 2000). The most common are C. annuum L., C. baccatum L., C. chinense
Jacq., C. frutescens L., and C. pubescens R. & P., and horticulturists have cultivated
these species worldwide for long time (Casali and Couto, 1984; Moscone et al., 2007). In
Brazil, the most indigenous is C. chinensis because this species has a large genetic diversity in the upper Amazon, and it is well adapted to the diverse environmental conditions
around the country (Lannes et al., 2007).
Pungency and flavor are fruit attributes of C. chinensis (Bosland, 1993) because of capsaicin and dihydrocapsaicin, which are components of an alkaloid complex responsible for 90% of the intense organoleptic sensation of heat (Kawada et al.,
1985). These fruits are also sources of vitamins A, complex B 1 and B 2, and minerals
such as dietary calcium, iron and phosphorus (Bosland, 1992). The content of vitamin C in the Capsicum fruit is higher than in Citrus. Furthermore, this species has
enormous potential as bedding plants in the ornamental agri-industry because the C.
chinensis fruit has large variability in size, shape and color (Stommel and Bosland,
2006; Segatto, 2007).
Mature fruit are very important in the agri-industry of dyes, and in numerous processes of the food industry of condiments. They are components of numerous dishes, and
may be eaten raw (Souza and Casali, 1984; Bosland, 1993). The agri-industry of dehydrated
peppers requires fruit with high dry matter and total soluble solids, thick pericarp and intense color, which are the foremost components for optimizing industrial yields (Casali and
Stringheta, 1984). In the raw form, consumers prefer the large and conical fruit that show
intense color, brilliant cuticle, smooth pericarp, and firm pulp (Casali et al., 1984). The
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
F.L. Finger et al.
1854
development of new cultivars may expand the Capsicum acreage to supply the ornamental
market with bedding plants and the pharmaceutical agri-industry with medicinal plants.
However, the success of breeding programs depends on genetic variability for selecting the
genotypes with the best commercial characteristics.
The Horticulture Gene Bank of the Universidade Federal de Viçosa (BGH-UFV)
has about 100 accessions of C. chinensis, in which Teixeira (1996) was the first to detect
genetic diversity for fruit size, morphology and color. Next, Lannes et al. (2007) detected
large genetic variability for capsaicinoid production. Finally, Pereira (2007) and Schuelter et al. (2010) reported the importance of the genetic factors to determine the fruit
characteristics of some accessions. However, more information is still required about
the physicochemical traits of these fruit, multi-categorical variables, and random amplified polymorphic DNA (RAPD) molecular markers to find these accessions suitable for
breeding programs. In addition, these same data may be used to permit the long-term
conservation of C. chinensis genotypes in gene banks.
Genetic resources of crop and wild species have been used as the sources of variability for morphological and agronomic traits (Carvalho et al., 2003). Analysis of genetic
diversity using quantitative or predictive methods has been used in the composition of
populations and lines capable of attending commercial purposes. Multiple data from every
accession and their evaluation by predictive analyses are expressed by dissimilarity indices, which represent the diversity of several groups (Cruz and Carneiro, 2003). Recently,
this divergence was evaluated by individual analyses of genetic distance and combined
analyses (Gonçalves et al., 2008; Kumar et al., 2009). Furthermore, several studies have
evaluated the genetic diversity of morphological and agronomic traits (Teixeira, 1996;
Oliveira et al., 1999; Rêgo et al., 2003) and molecular markers of Capsicum (Paran et al.,
1998; Buso et al., 2003; Toquica et al., 2003).
Thus, the objective of the present study was to document the characteristics of 49 accessions of C. chinensis stored in the BGH-UFV using RAPD markers, to estimate the genetic
diversity of physicochemical fruit traits, and to evaluate multi-categorical variables through
multivariate analyses.
MATERIAL AND METHODS
Material and crop conditions
Forty-nine accessions of C. chinensis collected in different Brazilian regions and
stored in the BGH-UFV (Table 1) were selected based on their fruit appearance and plant
architecture. Their physicochemical fruit traits and the multi-categorical variables were evaluated in the Department of Plant Production, Viçosa, Minas Gerais State, Brazil. The RAPD
analyses, otherwise, were carried out in the Biotechnology Laboratory of the Universidade
Paranaense, Toledo, Paraná State, Brazil.
Pepper seeds germinated on Styrofoam trays filled with organic growing medium. The
45-day-old bedding plants were transplanted to the field in a completely randomized block
design with three replications. Every replication had the plants spaced 0.5 m within rows and
1.0 m between rows, where the morphological and agronomic traits were evaluated from the
inner four in six plants.
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
Genetic diversity of Capsicum chinensis
1855
Table 1. Origin of 49 accessions of Capsicum chinensis stored in the BGH-UFV and their characteristics
evaluated by qualitative multi-categorical traits.
Accession
OriginMulticategorical traits
BGH 1694-5
BGH 1694-6
BGH 1694-7
BGH 1714-9
BGH 1714-11
BGH 1716-14
BGH 1716-16
BGH 1716-17
BGH 1716-18
BGH 1716-19
BGH 1723-22
BGH 1724-23
BGH 1747-26
BGH 1747-27
BGH 4199-30
BGH 4201-32
BGH 4213-34
BGH 4223-39
BGH 4285-40
BGH 4289-44
BGH 4289-45
BGH 4355-46
BGH 4725-51
BGH 4731-53
BGH 4733-54
BGH 4733-55
BGH 4733-56
BGH 4744-57
BGH 4750-59
BGH 4756-67
BGH 4756-70
BGH 4756-71
BGH 5012-72
BGH 5012-76
BGH 6009-78
BGH 6228-79
BGH 6228-82
BGH 6233-83
BGH 6233-84
BGH 6233-85
BGH 6239-86
BGH 6344-87
BGH 6369-90
BGH 6371-93
BGH 6371-94
BGH 6371-95
BGH 6378-98
BGH 6387-100
BGH 6515-101
GH SP SC CC FS IFC MFCSFT
Cuiabá, MT
Cuiabá, MT
Cuiabá, MT
Pindaré-Mirim, MA
Pindaré-Mirim, MA
Pindaré-Mirim, MA
Pindaré-Mirim, MA
Pindaré-Mirim, MA
Pindaré-Mirim, MA
Pindaré-Mirim, MA
Santarém, AM
Santarém, AM
Ituiutaba, MG
Ituiutaba, MG
Belém, PA
Belém, PA
Tefé, AM
Campinas, SP
Rondonópolis, MT
Rondonópolis, MT
Rondonópolis, MT
Pedro Afonso, GO
Urucará, AM
Autazes-Mirim, AM
Manaus, AM
Manaus, AM
Manaus, AM
Canabuoca-Manacapuru, AM
Joari-Itacotiara, AM
Joari-Itacotiara, AM
Joari-Itacotiara, AM
Joari-Itacotiara, AM
Água Branca, AL
Água Branca, AL
Belém, PA
Brasília, DF
Brasília, DF
Brasília, DF
Brasília, DF
Brasília, DF
Brasília, DF
Dourados, MS
Vila Nova-Anamam, AM
Vila Nova-Anamam, AM
Vila Nova-Anamam, AM
Vila Nova-Anamam, AM
Boca do Janacanan-Codajas, AM
Salvador, BA
Viçosa, MG
5
1
1
5 31
5 51
7 31
5 31
3 31
7 31
5 31
5 31
5 31
7 52
7 31
7 31
5 51
5 71
7 32
7 51
5 52
7 71
7 31
5 31
7 51
5 51
7 31
7 52
5 51
7 51
5 32
7 31
5 51
5 31
7 31
5 51
7 31
5 31
5 31
7 51
5 51
7 31
5 31
7 31
5 32
5 31
5 31
5 31
7 12
5
3
1
7 32
7 12
1
6
1
16 1
16 1
14 1
15 1
11 1
14 2
11 1
11 1
11 1
15 1
11 1
13 1
14 1
15 1
14 1
11 5
11 1
14 1
11 1
11 1
16 1
14 1
14 1
11 1
11 1
11 1
15 1
11 1
11 1
11 1
11 2
11 2
11 2
15 1
11 1
14 1
11 1
14 1
14 1
11 1
12 5
14 1
21 1
21 1
16 1
1
4
1
16 1
12 1
3
7
4 7
4 7
3 3
4 3
3 3
2 3
2 3
2 3
3 3
3 3
4 3
4 3
4 3
4 3
3 3
3 7
4 3
4 3
4 3
3 3
3 5
3 3
3 3
4 3
4 3
4 3
4 3
4 3
4 3
4 3
4 3
3 3
3 3
4 3
3 3
4 3
4 3
4 3
4 3
4 3
3 5
4 7
4 3
4 3
4 3
4
3
4 7
3 3
GH = growth habit; SP = stem pubescence; SC = stem color; CC = corolla color; FS = fruit shape; IFC = immature
fruit color; MFC = mature fruit color; SFT = shape of the fruit tip.
Morphological and agronomic traits
In five mature fruits, fruit length and width, pulp thickness, total fresh biomass, total dry
biomass, percentage of dry biomass, content of soluble solids, content of vitamin C, color intensity,
capsaicin, and dihydrocapsaicin using the method of Maillard et al. (1997), and the sum of both
capsaicinoids were all evaluated in the Post-Harvest Laboratory at the Department of Plant Production in the Universidade Federal de Viçosa. Vitamin C was determined using Tillmans’ method
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
F.L. Finger et al.
1856
(Instituto Adolfo Lutz, 1985), and color intensity was evaluated using the ASTA method 20-1.
Twelve plants of every accession were examined following the characteristics: prostrate (3), compact (5), or erect (7) growth habit; absent (1), scarce (3), moderate (5), or abundant (7) stem pubescence; elongated (1), triangular (2), rounded (3), conical (4), campanular
(5), or bell (6) fruit shape; pointed (3), blunt-tipped and absent (5) or sunken (7) fruit apex;
green (1) or purple (2) stem color; white (1), white-greenish (2) or green (2) corolla; green
(1), yellow (2), orange (3), red (4), brown (6), or black (7) immature fruit, and green (1),
yellow (2), orange (3), red (4), purple (5), brown (6), or black (7) mature fruit. These eight
multi-categorical traits were evaluated using the keys of minimum descriptors available at the
International Plant Genetic Resource Institute (IPGRI, 1995).
RAPD molecular markers
Six-day-old leaf blades were collected from bedding plants to characterize the accessions
using the RAPD markers. The protocol for DNA extraction and purification was described by
Fulton et al. (1995), and the amplifications were carried out on the Thermo PX2 thermocycler (Williams et al., 1990). The amplified fragments were submitted to horizontal electrophoresis on 1.2%
agarose gel, stained with 2 mg/L ethidium bromide and revealed by a photo documentation system.
Statistical analyses
Data of physicochemical traits were submitted to the analysis of variance to determine
genetic variability, and the Scott-Knott test (1974) was used to cluster the accessions. Estimates of
genetic and environmental parameters were used to evaluate the breeding chances of these traits.
The presence of multicollinearity in the correlation matrix was investigated before the
multivariate analyses to eliminate any bias in the estimates of genetic parameters. Afterward,
the generalized Mahalanobis distance was used to determine the dissimilarity matrix for the
quantitative traits, Cole-Rodgers et al.’ coefficient (1997) for the multi-categorical traits, and
Nei and Li’s coefficient (1979) for the molecular markers. Next, the Mantel test (1967) was
applied to determine the association between the genetic distance matrices.
The clusters were formed by optimizing Tocher’s method (Rao, 1952) through the different genetic distance matrices; the relative contribution of every trait to genetic divergence
was estimated by Singh’s method (1981). The analyses of canonical variables to eliminate
those physicochemical traits with no influence on Tocher’s clustering were suggested by Rêgo
et al. (2003). All analyses were carried out using the GENES software (Cruz, 2006a,b).
RESULTS AND DISCUSSION
All responses from the physicochemical traits (Table 2) were significantly different (P <
0.01) and showed evidence of genetic variability between the genotypes. Data of fruit length and
width, pulp thickness, capsaicin, dihydrocapsaicin, total capsaicinoids, soluble solids, and vitamin
C had the coefficients of variation between 5.30 and 20.44%, which accounted for accuracy in the
experimental errors. Otherwise, the magnitude of these coefficients from data for fresh and dry
biomass and fruit color indicated discrepancy among the accessions. Usually, as these traits are affected by the environment, there are difficulties in controlling them under experimental conditions.
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
d.f. LengthWidth
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
0.01
0.30*
0.04
0.65
33.60
84.10
1.33
0.08
0.10
0.01
5.87
29.98
93.26
2.14
14.33
15.37
1.03
TDB
0.98
46.11*
3.10
TFB
13.06
20.44
77.87
1.08
8.37
10.74
2.38
2.04
32.23*
7.13
%TDB
3.88
5.30
99.83
14.13
8.45
8.47
0.01
0.18
25.41*
0.04
CAP
1.00
7.90
99.68
10.25
0.66
0.67
0.01
4 x 10-3
95.65*
0.60
DHCAP
4.88
5.34
99.82
13.83
13.04
13.06
0.02
0.19
39.20*
0.68
TCAP
8.11
13.43
77.99
1.08
1.40
1.80
0.40
0.34
5.40*
1.19
CSS
97.65
17.32
78.22
1.09
342.90
438.33
95.43
20.20
1315.0*
286.31
Vit C
142.91
36.82
91.65
1.91
10142.8
11066.0
0.002
4960.78
33198.1*
2769.69
ECI
0.21
17.38
82.87
1.27
2 x 10-3
2 x 10-3
4 x 10-4
2 x 10-4
7 x 10-3*
1.3 x 10-3
PT
d.f. = degrees of freedom. Length and width are reported in mm. TFB = total fresh biomass (in gram); TDB = total dry biomass (in gram); %TDB = percentage
of dry biomass; CAP = capsaicin (in mg/g); DHCAP = dihydrocapsaicin (in mg/g); TCAP = total of capsaicin (in mg/g); CSS = content of soluble solids (in
°Brix); Vit C = vitamin C (in mg/100 g fresh fruit); ECI = extractable color intensity (in ASTA units); PT = pericarp thickness (in mm). , ,
= genotypic,
phenotypic and environmental variances, respectively. *Significant at 1% by the F-test.
Source of variation
Block
2
50.77
6.61
Accessions
48
761.47* 187.58*
Error
96
59.90
7.89
Genetic and
environmental parameters
General means
41.14
20.87
CVe (%)
18.81
13.45
H 2
92.13
95.79
CVg:CVe
1.97
2.75
233.85
59.89
253.82
62.52
19.96
2.62
Table 2. Analyses of variance, means, coefficients of environmental variation (CVe) and genotype (CVg) determination (H2), CVg to CVe ratios, and genotypic,
phenotypic and environmental variances of 12 physicochemical fruit traits evaluated in 49 accessions of Capsicum chinensis.
Genetic diversity of Capsicum chinensis
1857
©FUNPEC-RP www.funpecrp.com.br
F.L. Finger et al.
1858
The importance of genetic factors on these morphological and agronomic traits is indicated by the high genotypic coefficients of determination between 77.87 and 99.83%. Therefore,
fruit length and width, fresh biomass, capsaicin, dihydrocapsaicin, total capsaicinoids, dry biomass, pericarp thickness, and extractable color can be selected by the mass method, because the
genotypic coefficients of determination were above 80%, the ratio CVg to Cve was higher than
one, and there was high genetic variance. Otherwise, breeding methods capable of controlling the
environment must be used to select plants with high percentage of dry biomass, total soluble solids and vitamin C, because the coefficient of determination from the genotypes was close to one.
The Scott-Knott test (P < 0.05) allowed the clustering of these accessions into five
groups based on the fruit length between 14.15 and 76.22 mm. The accessions BGH 6371-93,
BGH 4733-54, BGH 4223-39, and BGH 4285-40 are clustered into group A because of their
long fruit, unlike BGH 6515-101 and BGH 4289-44, which are clustered into an opposite
group because of their short fruit. Otherwise, all the accessions are clustered into six groups
based on fruit width. The cluster A has just BGH 6009-78 with an estimate of 42.89 mm, but
the group B contains the accessions BGH 4355-46, BGH 4731-53, BGH 6233-84, and BGH
6371-95. Eighty-seven percent of the accessions, however, have fruit width less than 30 mm.
The ratio fruit length to fruit width ranges between 0.68 and 6.61, and long fruit is the predominant shape. The BGH 6371-94 accession shows the highest ratio.
Pulp thickness ranges from 1.0 to 3.5 mm, and all the accessions were clustered into
only three groups in which BGH 6009-78, with the thickest pulp, was clustered together with
BGH 4223-39, BGH 6233-85, BGH 6378-98, BGH 6369-90, and BGH 1694-07; all of them
have about 3.0 mm of pulp. Other accessions were clustered into group B (28.6%) because
of their moderate pulp thickness and C (59.2%) because of their thinner pulps. The accession
BGH 6371-94 had 1.17 mm thickness and BGH 4733-56 had 1.11 mm thickness, representing
the thinnest fruit pulp (Table 3).
The total biomass is between 0.99 and 19.15 g. In the cluster A, BGH 6009-78 is the
only accession with the highest estimate (P < 0.05), and it is followed by the cluster B with
BGH 6233-84, BGH 4223-39, BGH 4731-53, and BGH 6233-85. Ninety percent of the accessions, otherwise, have estimates below 11 g, and they were clustered into C, D and E.
Fresh biomass is predominantly lower among the accessions (Table 3). High fruit dehydration accounted for the dry biomass between 0.17 and 1.42 g, and only four clusters were
formed using the Scott-Knott test (P < 0.05). BGH 6009-78, BGH 4223-39 and BGH 4731-53
with high fresh biomass together with BGH 6371-93 were clustered into group A. Otherwise,
BGH 6233-84 and BGH 6233-85, which were clustered into A for total biomass, were here
clustered into B. This fact raises the importance of the water content because of the large
variation, between 7.05 and 20.92%, in the percentage of dry biomass in these genotypes.
The accessions were distributed into the groups with high (A) and low (B) percentage of dry
biomass (P < 0.05). Cluster A with 31% of the accessions had the minimum dry biomass equal
to 14.42%, although BGH 4733-56 and BGH 4289-44 had percentages of about 21% despite
the low level of total fresh biomass in these small fruit. Similar responses were reported by
Lannes et al. (2007), who found a negative correlation between these traits. Nevertheless, both
traits have sufficient genetic variability to participate in the genotype selection to supply the
fresh fruit market or the food industry with new cultivars.
The content of total soluble solids ranges between 5.93 and 12.90°Brix, and all the
accessions were clustered into only three groups. BGH 4733-56, with the highest content, was
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
Genetic diversity of Capsicum chinensis
1859
Table 3. Means of 12 physicochemical fruit characteristics of 49 accessions of Capsicum chinensis in the
Horticulture Gene Bank at the Universidade Federal de Viçosa, MG, Brazil.
Accession
LengthWidth
TFB
TDB
%TDB
CSS
Vit C
ECI
PT
BGH 1694-5
BGH 1694-6
BGH 1694-7
BGH 1714-9
BGH 1714-11
BGH 1716-14
BGH 1716-16
BGH 1716-17
BGH 1716-18
BGH 1716-19
BGH 1723-22
BGH 1724-23
BGH 1747-26
BGH 1747-27
BGH 4199-30
BGH 4201-32
BGH 4213-34
BGH 4223-39
BGH 4285-40
BGH 4289-44
BGH 4289-45
BGH 4355-46
BGH 4725-51
BGH 4731-53
BGH 4733-54
BGH 4733-55
BGH 4733-56
BGH 4744-57
BGH 4750-59
BGH 4756-67
BGH 4756-70
BGH 4756-71
BGH 5012-72
BGH 5012-76
BGH 6009-78
BGH 6228-79
BGH 6228-82
BGH 6233-83
BGH 6233-84
BGH 6233-85
BGH 6239-86
BGH 6344-87
BGH 6369-90
BGH 6371-93
BGH 6371-94
BGH 6371-95
BGH 6378-98
BGH 6387-100
BGH 6515-101
27.78e
22.17d
23.43e
21.27d
33.40d
28.55c
41.82d
21.99d
40.32d
25.46c
36.81d
19.09d
37.74d
20.77d
52.29c
13.73e
54.69c
17.55d
62.83b
14.17e
29.56e
22.32d
21.71e
12.87e
16.57e
15.27e
37.17d
19.53d
39.47d
26.71c
48.33c
17.91d
25.39e
18.58d
70.43a
22.98d
70.27a
24.81c
19.41e
10.89f
40.13d
11.43f
27.27e
34.68b
33.92d
20.07d
58.17b
33.95b
74.16a
20.02d
61.03b20.52d
41.35d10.66f
42.81d26.27c
26.69e
14.03e
36.23d
15.07e
23.09e
12.25f
35.85d
15.99e
25.66e
10.37f
38.75d
10.58f
38.31d
42.89a
47.01c
12.86e
37.07d
24.26c
61.53b
21.16d
50.31c
37.38b
56.94b
30.55c
62.19b
22.73d
23.74e
14.89e
32.78d
32.95b
76.22a
23.69c
57.57b
8.69f
21.67e
31.83b
48.74c
25.89c
33.19d
28.77c
14.15e
11.70f
3.55e
3.18e
9.65c
5.92d
5.58d
4.28d
6.06d
4.28d
5.80d
5.29d
4.81d
1.65e
1.38e
4.22d
6.96d
5.67d
2.88e
13.08b
9.50c
0.99e
2.42e
6.77d
4.24d
12.46b
6.65d
6.21d
1.86e
7.47d
2.46e
2.90e
1.34e
4.21d
1.26e
1.93e
19.15a
3.02e
5.90d
8.50c
13.28b
12.40b
7.65d
2.40e
11.00c
9.38c
2.25e
6.90d
10.58c
7.59d
1.02e
0.50c
0.41d
0.98b
0.86b
0.62c
0.49c
0.72c
0.52c
0.66c
0.57c
0.50c
0.23d
0.25d
0.54c
0.68c
0.75b
0.45c
1.42a
0.95b
0.20d
0.38d
0.64c
0.55c
1.17a
0.86b
0.70c
0.33d
0.85b
0.42d
0.38d
0.17d
0.46c
0.22d
0.28d
1.38a
0.52c
0.65c
0.86b
1.05b
0.95b
0.90b
0.45c
0.93b
1.33a
0.38d
0.65c
1.03b
0.83b
0.17d
13.90b
12.91b
10.21b
16.73a
10.79b
11.68b
12.45b
12.51b
11.83b
11.63b
10.45b
14.72a
18.27a
13.68b
10.40b
12.74b
16.05a
10.91b
10.34b
20.53a
15.53a
10.56b
13.66b
9.76b
12.83b
11.65b
20.92a
11.80b
17.43a
13.78b
13.26b
10.85b
17.64a
14.42a
7.05b
17.04a
11.99b
10.68b
8.27b
8.06b
13.02b
19.13a
9.29b
14.52a
17.46a
9.56b
9.37b
10.91b
16.91a
8.36c
8.49c
6.48c
8.90b
7.20c
8.59b
9.44b
8.29c
8.80b
8.30c
7.75c
8.26c
7.64c
8.81b
7.19c
9.09b
7.88c
7.03c
6.59c
10.86a
11.74a
6.75c
7.46c
7.20c
8.09c
7.84c
12.90a
7.47c
9.31b
7.05c
9.08b
7.86c
9.69b
8.85b
6.76c
7.99c
7.80c
7.20c
7.04c
5.93c
5.97c
8.69b
6.20c
9.18b
8.47c
7.80c
7.17c
8.30c
8.04c
98.65a
81.41b
83.65b
91.65b
86.42b
80.95b
86.31b
101.31a
106.15a
100.92a
99.47a
78.60b
118.94a
97.48a
107.57a
113.15a
112.05a
100.07a
111.66a
76.01b
119.02a
56.17c
113.40a
81.62b
134.30a
103.66a
105.47a
91.89b
115.96a
127.01a
135.22a
108.25a
97.12a
117.98a
73.98b
138.85a
110.09a
76.54b
116.07a
101.31a
70.60b
112.90a
86.81b
51.38c
38.76c
89.41b
74.91b
96.55a
107.57a
88.78e
229.08c
123.97d
105.54e
181.92d
56.66e
30.35e
75.25e
56.75e
60.21e
130.93d
234.57c
383.74b
273.77c
140.49d
38.96e
111.50e
133.86d
174.51d
34.87e
80.88e
50.70e
62.82e
54.66e
185.39d
158.30d
595.84a
188.45d
315.24b
99.61e
72.32e
171.16d
49.28e
65.91e
87.49e
109.16e
328.74b
113.11e
154.64d
214.38d
137.93d
44.50 e
97.25e
152.53d
286.36c
128.47d
142.16d
81.36e
108.72e
0.18c
0.17c
0.28a
0.24b
0.22b
0.20c
0.25b
0.20c
0.19c
0.16c
0.19c
0.16c
0.19c
0.18c
0.24b
0.24b
0.20c
0.30a
0.20c
0.14c
0.15c
0.17c
0.25b
0.24b
0.15c
0.18c
0.11c
0.23b
0.17c
0.18c
0.19c
0.23b
0.14c
0.15c
0.36a
0.16c
0.23b
0.22b
0.24b
0.30a
0.17c
0.20c
0.29a
0.20c
0.12c
0.24b
0.29a
0.22b
0.17c
CAP DHCAP
TCAP
3.91k 0.62k
4.53l
3.89k 0.81j
4.70l
1.42o 0.45l
1.87q
1.70o 0.44l
2.14p
1.89n 0.47l
2.36p
0.00q 0.00m
0.00s
0.00q 0.00m
0.00s
10.32a 3.30a
13.62a
5.58h 1.18h 6.76i
1.45o 0.40l
1.85q
3.13m 1.78e
4.91k
7.15f 1.35g
8.50f
2.80m 0.56l
3.36n
3.84k 0.10i
4.84l
4.04k 1.00i
5.04k
0.00q 0.00m
0.00s
4.86i 1.76e
6.62i
0.00q 0.00m
0.00s
3.99k 1.07i
5.06k
4.71i 3.42a
8.13f
4.82i 1.36g
6.18j
4.94i 1.05i 5.99j
3.69k 1.03i
4.72l
0.00q 0.00m
0.00s
7.48e 2.34b
9.82d
4.88i 1.17h 6.05j
6.97f 1.70f
8.67f
0.00q 0.02m
0.03s
6.84f 2.16c
9.00e
6.72f 1.58f
8.30f
6.28g 1.11h
7.39h
3.86k 0.84j
4.70l
8.30d 1.23h
9.53d
6.88f 1.05i
7.93g
0.00q 0.00m
0.00s
9.42b 1.96d
11.38b
7.76e 1.60f
9.35d
0.64p 0.08m
0.72r
3.42l 0.66k
4.08m
0.00q 0.00m
0.00s
4.37j 1.44g
5.81j
2.18n 0.49l
2.67o
2.06n 0.79j
2.85o
0.24q 0.46l
0.70r
2.95m 0.99i
3.94m
7.58e 1.88d
9.46d
0.00q 0.00m
0.00s
4.42j 0.79j
5.21k
8.81c 1.85d
10.66c
For abbreviations, see legend to Table 2. *Significant at 5% by the Scott-Knott test. Means followed by the same
letter in each columm belong to a homogeneous group at the 5% probability (Scott-Knott, 1974).
clustered into A together with BGH 4289-44 and BGH 4289-45 with levels above 10.87°Brix.
Next, 24.90% of the accessions were clustered into group B because of their intermediate levels between 8.59 and 9.69°Brix. Finally, 69.39% of the accessions were clustered into group
C because their totals were lower than 8.49°Brix (Table 3).
Vitamin C contents between 38.75 and 138.84 mg/100 g fresh pulp allowed for clustering 61% of the accessions into group A, in which all of them have vitamin C concentrations
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
F.L. Finger et al.
1860
higher than 95 mg. The averages of BGH 6228-79, BGH 4756-70 and BGH 4733-54 are all
higher than 130 mg/100 g fresh pulp (Table 3).
Color intensity ranged from 30.35 to 595.84 ASTA units, and some of these values are
higher than the estimates reported by Gomez et al. (1998), who investigated commercial varieties of C. chinensis under greenhouse cultivation. These authors found values between 190
and 360 units under field conditions, where their values ranged between 181 and 335 units.
These accessions were clustered into five groups in which A had only the BGH 4733-56 because of its intensive red color followed by cluster B with BGH 1747-26, BGH 6228-82, and
BGH 4750-59. In summary, there is genetic variability for fruit color intensity, which depends
on the pigment type, although red and orange have the highest magnitudes.
Total capsaicin ranges from trace levels to 13.62 mg/g dry biomass, and these levels are
similar to the data from Gibbs and O’Garro (2004), who reported estimates between 0.25 and 16.55
mg/g dry biomass. Total capsaicin showed the highest genetic diversity (P < 0.05), which accounted
for the distribution of the accessions among 18 clusters. Accessions with the highest content of
capsaicin, such as BGH 1716-17, were clustered into group A, BGH 6228-79 into B, BGH 6515101 into C, BGH 5012-72 into D, and BGH 6228-82 into E. Other accessions had intermediate
estimates, except those with sweet fruit, such as BGH 1716-14, BGH 1716-16, BGH 4201-32, BGH
4223-39, BGH 4731-53, BGH 6009-78, BGH 6233-85, and BGH 6378-98 (Table 3).
The genetic variability of capsaicin ranges from 0.00 to 10.32 mg/g dry biomass, and
17 clusters were formed (P < 0.05). The dihydrocapsaicin, otherwise, ranges from 0.00 to 3.42
mg, but the accessions were clustered into 13 groups. All of them have higher levels of capsaicin than dihydrocapsaicin, except BGH 6371-93 (Table 3). According to Bosland (1993), the
different combinations of capsaicinoids induce individual characteristics of fruit pungency,
although both capsaicinoids account for the same taste of heat.
The multi-categorical descriptors (IPGRI, 1995) showed genetic variability (Table
1). The growth of most accessions (53.06%) is classified as compact, and they are followed
by those with erect growth (44.89%). Just BGH 1716-14 grows prostrated on the ground.
Stems with scarce pubescence are found in 61.22% of the accessions followed by those with
intermediate (28.57%), smooth (6.12%) and abundant hairs (4.08%). In terms of stems, green
is the predominant color (81.63%), but in the other accessions they are purple. The color of
the corolla is whitish in 95.91% of the accessions and white-greenish in the others. In immature fruit, the predominant color is green (87.75%), followed by yellow (8.16%) and purple
(4.08%). Mature fruit are red (61.22%), orange (32.65%) and yellow (6.12%).
In terms of shape (Table 1), there is a high percentage of accessions with longitudinal
fruit (47.0%) followed by those with the conical fruit (25.0%). The other accessions have campanular (10.0%), bell (12.0%), triangular (4.0%), and rounded fruit (2%). Most of the accessions have pointed fruit tips (83.67%), but others have no tips (4.08%) or sunken tips (12.25%).
Thirteen RAPD primers (41.0%) in 32 are polymorphic. Forty-four bands (15.4%) are
polymorphic for the identified loci in 286 amplifications. Similar results were reported by Paran
et al. (1998), who studied the genetic diversity of C. annuum and found 20.0%, and Rodriguez
et al. (1999), who studied numerous species in the genus Capsicum, also found the same 20%.
The variance and covariance matrices detected severe multicollinearity. Therefore, the contents of capsaicin and dihydrocapsaicin were eliminated before the multivariate analysis because
of their high correlations with total capsaicin, and their substantial contribution to multicollinearity.
Furthermore, the generalized Mahalanobis distance evaluated the redundancy in the physicochemiGenetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
Genetic diversity of Capsicum chinensis
1861
cal traits from which only the total dry biomass was eliminated because of the autovalue in the last
autovector detected by the canonical variables. In terms of the relative contribution of physicochemical traits, total capsaicin accounts for 88.88% of the variation followed by fruit width (2.65%), fruit
length (2.10%), fruit dry matter (1.85%), extractable fruit color (1.76%), vitamin C (1.14%), soluble
solids (0.87%), pulp thickness (0.73), and percentage of fruit dry matter (0.22%). These results accounted for the variation in total capsaicin, which contributed to the genetic divergence.
Tocher’s method, cited by Rao (1952), using Mahalanobis’ genetic distance (1936),
clustered all the accessions into nine groups (Table 4). The cluster I consists of all the accessions with sweet fruit, including BGH 1716-14, BGH 1716-16, BGH 4201-32, BGH 422339, BGH 4731-53, BGH 4744-57, BGH 6233-85, and BGH 6378-98, except BGH 6009-78,
which is clustered into IX. All accessions in the clusters II, III, V, VI, VII, and VII contain
concentrations of total capsaicin from average to high levels, but the most pungent accessions
such as BGH 1716-17 is clustered into VI, BGH 4733-54 and BGH 6228-79 are clustered into
V, and BGH 6515-101 is clustered into group III. Otherwise, BGH 1747-26, BGH 6344-87
and BGH 6369-90 are clustered into group IV because of the total capsaicin from low to averTable 4. Clustering of 49 accessions of Capsicum chinensis from the BGH-UFV by Tocher’s method, according
to physicochemical characteristics of fruit, multi-categorical traits, RAPD markers, and combined analysis.
CharacteristicsCluster Accessions
PhysicochemicalI
7*, 9, 11, 14, 16, 19, 32, 39, 53, 57, 83, 85, 93, 98
II
5, 6, 18, 22, 27, 30, 34, 40, 45, 46, 51, 55, 71 84, 86, 100
III
23, 44, 59, 67, 70, 72, 76, 82, 95, 101
IV
26, 87, 90
V
54, 79
VI17
VII56
VIII94
IX
78
Multi-categorical
I
6, 7, 9, 11, 14, 16, 17, 18, 19, 23, 26, 27, 30, 32, 39, 40, 44, 45, 51, 53, 54, 55,
56, 57, 59, 67, 70, 71, 72, 76, 78, 79, 82, 83, 84, 85, 86, 90, 93, 94, 95, 98
II
5, 22, 34, 46, 101
III100
IV87
RAPD
I
5, 6, 7, 14, 16, 18, 19, 22, 23, 26, 27, 30, 32, 34, 39, 40, 45, 46, 51, 53,
54, 55, 56, 57, 67, 70, 71, 76, 78, 79, 82, 83, 84, 85, 87, 95, 98, 101
II
93, 94
III
9, 11, 59
IV
86, 100
V72
VI90
VII44
VIII17
Combined analysis
I
5, 6, 14, 16, 17, 18, 19, 32, 45, 51, 55, 67, 70, 71, 78, 82, 83, 85
II
93, 94
III
7, 9, 11, 84
IV
26, 27, 30, 39
V
22, 56, 59
VI
40, 54
VII
46, 72, 76, 79
VIII
90, 98
IX
34, 87, 95, 101
X
57, 100
XI
23
XII
86
XIII
53
XIV
44
*In this table, the accessions are represented by their two final numbers.
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
F.L. Finger et al.
1862
age levels, and BGH 6371-94 is clustered alone into group VIII.
Tocher’s method based on the matrices of Cole-Rodgers et al. (1997) for dissimilarities
formed four clusters, and the Nei and Li (1979) matrices formed eight (Table 4). Both methods
were poor in discriminating the accessions, and their results were found in disagreement with
the Mahalanobis distance. This disagreement is verified by the absence of significant differences detected by the Mantel test when the pairs of distance matrices are compared. Therefore,
the combined analysis was used to form 14 clusters by Tocher’s method after the sum of the
standardized distance with 36.7% of the accessions (Table 4). Significant correlations with the
Mahalanobis (r = 0.55), Cole-Rodgers (r = 0.59) and Nei and Li (r = 0.60) distance matrices
were detected by the Mantel test (10,000 simulations) to verify the association between these
matrices and the individual analyses. These results showed the equitable contribution of the
different traits to genetic diversity without the bias of the Gower algorithm (1971), because
of moderate contribution in the combined matrices, which depends on the number of traits.
Sweet fruit (BGH 1716-14, BGH 1716-16, BGH 4201-32, and BGH 6233-85) and
less pungent accessions such as BGH 1716-19 and BGH 6233-83 were clustered according
to physicochemical characteristics, and they maintained here the same constitution because
of their great genetic similarities. Otherwise, the accession with sweet fruit (BGH 6009-78),
which was alone in cluster IX, was then grouped into cluster I by the combined analysis.
In terms of origin, the use of geographic sites to discriminate the accessions revealed
inconsistency. BGH 6369-90, BGH 6371-93, BGH 6371-94, and BGH 6371-95 collected in Vila
Nova-Anaman, AM, were distributed in the cluster II (BGH 6371-93 and BGH 6371-94), VIII
(BGH 6369-90 and BGH 6378-98) and IX (BGH 6371-95). BGH 6371-95, together with the accessions collected in Tefé, AM (BGH 4213-34), Dourados, MS (BGH 6344-87) and Viçosa, MG
(BGH 6515-101) consolidated the cluster IX. The accessions collected in Pindaré-Mirim, AM,
BGH 1716-14, BGH 1716-16, BGH 1716-17, BGH 1716-18, and BGH 1716-19 were clustered
into I, and BGH 1716-9 and BGH 1716-11 were clustered into III together with BGH1694-7 and
BGH 6233-84, collected in Cuiabá, MT, and Brasília, DF, respectively. Other accessions from
Brasília, DF, were clustered into I (BGH 6228-82, BGH 6233-83 and BGH 6233-85), but the
cluster VII had the BGH 6228-79 and XII had the BGH 6239-86, only. The above accessions
and clusters showed no relationship between genetic diversity and geographic distance. According to Murty and Arunachalan (1966) and Upadhaly and Murty (1970), both cited by Cruz and
Carneiro (2003), genetic drift and plant selection in different environments can have more significant effects on diversity than the geographical location. Nevertheless, genetic contamination
during seed multiplication can affect these responses, because cross pollination in the Capsicum
genus ranges from 0.5 to 70% (Cruz and Carneiro, 2003), although the plants in the current study
were protected by white cloth to reduce the chances of contamination.
Regardless of the accession origin, clustering is essential for selecting divergent C.
chinensis genotypes. For example, within cluster I, which accounted for 36.7% of accessions,
it is possible to select accessions such as BGH 1716-17 to participate in breeding programs
because of its fruit pungency. Thus, this accession can be intercrossed, for example, with BGH
4223-39 and BGH 6233-84 from other clusters because of their fruit size.
Otherwise, as the analysis of genetic diversity is predictive because it does not consider the gene effects of traits, the crosses can produce progeny with unsatisfactory agronomic
responses. Nonetheless, the use of multivariate analyses to identify divergent accessions may
be essential for organizing the diallel permitting the evaluation of the types of gene effects and
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
Genetic diversity of Capsicum chinensis
1863
the crossing of pairs with better chances of success. Therefore, the most promising pungent accessions to constitute the diallel are BGH 1716-17, BGH 4289-44, BGH 4733-54, BGH 501272, BGH 6228-79, and BGH 6515-101. Otherwise, the better accessions with sweet fruit are
BGH 1716-16, BGH 4201-32, BGH 4223-39, BGH 4731-53, BGH 6233-85, and BGH 637898, of which BGH 1716-16, BGH 4201-32 and BGH 6233-85 were clustered into group I.
The genetic control of fruit pungency, which has been considered as dominant monogenic in most reports (Greenleaf, 1986; Stewart et al, 2007), has been easy to identify after
controlled crosses. Therefore, in-depth exploration of this genetic variability can be performed
using diallel crosses where the accessions have a different level of pungency. Finally, the formation of clusters to study the genetic control of pungency and sweetness using partial diallel
crosses may also be an interesting genetic strategy.
ACKNOWLEDGMENTS
Research supported by FAPEMIG, CAPES/DAAD and IPEAC/UNIPAR.
REFERENCES
Bosland PW (1992). Chiles: a diverse crop. Hort. Techn. 2: 6-10.
Bosland PW (1993). Breeding for quality Capsicum. Capsicum Eggplant Newsl. 12: 25-31.
Buso GSC, Amaral ZPS, Bianchetti LB, Machado FRB, et al. (2003). Genetic variability and phylogenetic analysis of
Brazilian species of Capsicum. Capsicum Eggplant Newsl. 22: 1-3.
Carvalho SIC, Bianchetti LB, Bustamante PG and Silva DB (2003). Catálogo de Germoplasma de Pimentas e Pimentões
(Capsicum spp) da Embrapa Hortaliças. Embrapa Hortaliças, Brasília.
Casali VWD and Couto FA (1984). Origem botânica de Capsicum. Inf. Agrop. 113: 8-10.
Casali VWD and Strigheta PC (1984). Melhoramento de pimentão e pimenta para fins industriais. Inf. Agrop. 10: 23-25.
Casali VWD, Pádua JG and Braz LT (1984). Melhoramento de pimentão e pimenta. Inf. Agrop. 10: 19-22.
Cole-Rodgers P, Smith DW and Bosland PW (1997). A novel statistical approach to analyze genetic resource evaluations
using Capsicum as an example. Crop Sci. 37: 1000-1002.
Cruz CD (2006a). Programa Genes: Estatística Experimental e Matrizes. Editora UFV, Viçosa.
Cruz CD (2006b). Programa Genes: Biometria. Editora UFV, Viçosa.
Cruz CD and Carneiro PCS (2003). Modelos Biométricos Aplicados ao Melhoramento Genético. Editora da UFV, Viçosa.
Fulton TM, Chunwongse J and Tanksley SD (1995). Microprep protocol for extraction of DNA from tomato and
herbaceous plants. Plant Mol. Biol. Rep. 13: 207-209.
Gibbs HAA and O’Garro LW (2004). Capsaicin content of West Indies hot pepper cultivars using colorimetric and
chromatographic techniques. Hort. Sci. 39: 132-135.
Gomez R, Pardo JE, Navarro F and Varón R (1998). Color differences in paprica pepper varieties (Capsicum annuum L.)
cultivated in a greenhouse and in the open air. J. Sci. Food Agric. 77: 268-272.
Gonçalves LSA, Rodrigues R, Amaral Junior AT and Karasawa M (2008). Comparison of multivariate statistical
algorithms to cluster tomato heirloom accessions. Genet. Mol. Res. 7: 1289-1297.
Gower JC (1971). A general coefficient of similarity and some of its properties. Biometrics 27: 857-874.
Greenleaf W (1986). Pepper Breeding. In: Breeding Vegetable Crops (Basset MJ, ed.). AVI, Connecticut, 69-134.
Instituto Adolfo Lutz (1985). Normas Analíticas do Instituto Adolfo Lutz; Métodos Químicos e Físicos para Análises de
Alimento. Instituto Adolfo Lutz, São Paulo.
IPGRI (International Plant Genetic Resources Institute) (1995). Descriptors for Capsicum (Capsicum spp). International
Plant Genetic Resources Institute, Rome.
Kawada T, Watanabe T, Katsura K, Takami H, et al. (1985). Formation and metabolism of pungent principle of Capsicum
fruits. XV. Microdetermination of capsaicin by high-performance liquid chromatography with electrochemical
detection. J. Chromatogr. 329: 99-105.
Kumar V, Sharma S, Sharma AK and Sharma S (2009). Comparative analysis of diversity base on morpho-agronomic
traits and microsatellite markers in common bean. Euphytica 170: 249-262.
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br
F.L. Finger et al.
1864
Lannes SD, Finger FL, Schuelter AR and Casali VWD (2007). Growth and quality of Brazilian accessions of Capsicum
chinense fruits. Scient. Hortic. 112: 266-270.
Mahalanobis PC (1936). On the generalized distance in statistics. Proc. Natl. Inst. Sci. India 2: 49-55.
Maillard MN, Giampaoli P and Richard HJ (1997). Analysis of eleven capsaicinoids by reversed-phase high performance
liquid chromatography. Flav. Fragr. J. 12: 409-413.
Mantel N (1967). The detection of disease clustering and a generalized regression approach. Cancer Res. 27: 209-220.
Moscone EA, Scaldaferro MA, Grabiele M and Cecchini NM, et al. (2007). The evolution of chilli peppers (Capsicum Solanaceae): a cytogenetic perspective. Acta Hort. 745: 137-170.
Nei M and Li WH (1979). Mathematical model for studying genetic variation in terms of restriction endonucleases. Proc.
Natl. Acad. Sci. U. S. A. 76: 5269-5273.
Oliveira VR, Casali VWD, Cruz CD and Pereira PR (1999). Avaliação da diversidade genética em pimentão através de
análise multivariada. Hortic. Bras. 17: 19-24.
Paran I, Aftergoot E and Shifriss C (1998). Variation in Capsicum annuum revealed by RAPD and AFLP markers.
Euphytica 99: 167-173.
Pereira GM (2007). Análise Dialélica e Pungência de Frutos em Capsicum chinense. Doctoral thesis, UFV, Viçosa.
Rao CR (1952). An Advanced Statistical Method in Biometric Research. John Wiley and Sons, New York.
Rêgo ER, Rêgo MM, Cruz CD and Cecon PR (2003). Genetic diversity analysis of peppers: a comparison of discarding
variable methods. Crop Breed. Appl. Biotech. 3: 19-26.
Reifschneider FJB (2000). Capsicum: Pimentas e Pimentões no Brasil. Embrapa Comunicação para Transferência de
Tecnologia. Embrapa Hortaliças, Brasília.
Rodriguez JM, Berke T, Engle L and Nienhuis J (1999). Variation among and within Capsicum species revealed by RAPD
markers. Theor. Appl. Genet. 99: 147-156.
Schuelter AR, Pereira GM, Amaral Junior AT and Casali VWD (2010). Genetic control of agronomically important traits
of pepper fruits analyzed by Hayman’s partial diallel cross scheme. Genet. Mol. Res. 9: 113-127.
Scott AT and Knott M (1974). A cluster analysis method for grouping means in the analysis of variance. Biometrics 30:
507-512.
Segatto FB (2007). Avaliação da Qualidade “Pós-Produção” de Pimenta Ornamental (Capsicum annuum) Cultivadas em
Vaso. Doctoral thesis, UFV, Viçosa.
Singh D (1981). The relative importance of characters affecting genetic divergence. Ind. J. Gen. Plant Breed. 41: 237-245.
Souza RJ and Casali WD (1984). Cultivares de pimentão e pimenta. Inf. Agrop. 113: 15-18.
Stewart C Jr, Mazourek M, Stellari GM, O’Connell M, et al. (2007). Genetic control of pungency in C. chinense via the
Pun1 locus. J. Exp. Bot. 58: 979-991.
Stommel JR and Bosland PW (2006). Ornamental Pepper, Capsicum annuum. In: Flower Breeding and Genetics: Issues,
Challenges and Opportunities for the 21st Century (Anderson N, ed.). Springer, Dordrecht, 561-599.
Teixeira R (1996). Diversidade em Capsicum: Análise Molecular, Morfoagronômica e Química. Master’s thesis, UFV,
Viçosa.
Toquica SP, Rodríguez F, Martinez E, Duque MC, et al. (2003). Molecular characterization by AFLPs of Capsicum
germplasm from the Amazon department in Colombia. Gen. Res. Crop Evol. 50: 639-647.
Williams JG, Kubelik AR, Livak KJ, Rafalski JA, et al. (1990). DNA polymorphisms amplified by arbitrary primers are
useful as genetic markers. Nucleic Acids Res. 18: 6531-6535.
Genetics and Molecular Research 9 (3): 1852-1864 (2010)
©FUNPEC-RP www.funpecrp.com.br